|
|
| Fig. 1: Overview of ADS Reactors. (1) Protons are first accelerated in a LINAC and then further accelerated in an isochronous cyclotron (2) The high energy protons then collide with a heavy metal spallation medium, generating dozens of neutrons (an incoming 1.5 GeV proton colliding with a LBE target will generate on average 20 neutrons per collision.) [4] (3) Primary neurons (red lines) from spallation collide with nuclei in the subcritical core (blue area) are then captured by fertile material (Th-232 or U-238) to breed additional fissile material or fissioning fissile material (U-233, U-235 and Pu-239), thus generating first-generation neutrons (orange lines). In turn, the first-generation neutrons collide with other nuclei in the core to be captured or cause fission, generating second-generation neutrons (green lines). In the T-EA design an average of 10 to 20 neutrons are produced per primary neutron generated from spallation. [1] (Image Source: F. Rabbat Neto) |
Accelerator Driven Systems (ADS) are a next-generation nuclear reactor technology aimed at incinerating nuclear waste and addressing supercriticality concerns with conventional nuclear reactors. The original designs operate by colliding a high energy proton against a heavy metal source (known as a spallation target) to generate neutrons that initiate and sustain fission in a subcritical core. [1,2] The external supply of neutrons from the spallation source allows the reactor to operate below criticality, where the number of neutrons produced by fission is less than the number of neutrons consumed in the core, significantly reducing the risk of supercritical catastrophes such as Chernobyl. The large buffer between criticality and the operating conditions of the reactor make accidental supercriticality far less likely than in a critical reactor that operates right under supercriticality. Beyond safer energy generation, the technology promises high versatility of fuel sources and could theoretically serve as a transuranic, minor actinide and long-lived fission fragment transmuter, significantly reducing the radioactive load and lifetime of high-level radioactive waste. [3]
Several designs for ADS have been proposed by different authors. The designs that have gotten the most recognition and will be the focus here are the two from Carlo Rubbia: the original Thermal Energy Amplifier (T-EA) and its subsequent iteration, the Fast Energy Amplifier (F-EA). [1,2] Both designs operate by combining a proton beam accelerator with a spallation target and a subcritical core. A high-energy and high-flux proton beam generated in a combined LINAC and cyclotron accelerator is directed to a spallation target. [1] Upon collision, the nuclei of the spallation target undergo an intranuclear cascade (series of direct reactions) resulting in the ejection of nucleons or small groups of nucleons from the target nucleus. [1] This occurs thorugh multi-fragmentation (breaking down to several smaller particles), evaporation (ejection of nucleons), or fission. [1] These neutrons then collide with the subcritical fuel surrounding the spallation target, prompting fission and neutron capture. [1] Both the T-EA and F-EA designs act as breeder reactors, meaning that more neutrons are captured for breeding fissile elements than are used for fission. [1,2] To kick start the subcritical reaction, both designs suggest an initial fuel loading with fissile material (U-233, U-235 or Pu-239) mixed in a matrix of fertile material (Th-232 or U-238) to sustain initial subcritical conditions. [1,2] Neutrons generated from spallation collide with fissile material to cause fission or are captured by Th-232 or U-238 to produce U-233 and Pu-239 respectively. These later serve as fission fuel. The fission of the fissile isotopes generates first-generation neutrons that collide with other nuclei in the fuel core, causing further fission or capture. [1]
Initial estimates from the T-EA design suggest that for every neutron generated from spallation 10 to 20 daughter neutrons are generated from fission in a core designed to operate at a criticality factor of 0.9 to 0.95 respectively. [1] Estimates from the F-EA suggest an even greater number of daughter neutrons are generated per incoming neutron. [2] The role of spallation is to maintain the reaction, as fission generates a finite number of neutrons and cannot be sustained without an external source in the ADS design. [1] Fig. 1 outlines how an ADS operates.
On waste incineration, ADS provides high flexibility of fuel elements, allowing the fuel to be spiked with transuranic elements, minor actinides and long-lived fission fragments (LLFFs). These can undergo fission or capture to be broken into other nonradioactive elements or daughter products with shorter half-lives. [3]
As each external neutron can generate 10 to 20 neutrons from fission, maximizing the number of external neutrons is a key maximizing the energetic output from an ADS system. However, generating these neutrons must be energetically cheap for the ADS to be competitive with critical nuclear reactors. The more energy required to produce these external neutrons, the higher the cost of operating the ADS and the lower its net energy output per fuel input. In both of Rubbia's designs proton spallation is chosen as the neutron source, as it minimizes the energetic costs per produced neutron compared to other neutron-generating nuclear reactions. [4]
While proton spallation may be the least energetically expensive source of external neutrons, within proton spallation there are several different choices of materials and beam energies with varying neutron yield. The neutron yield per incident proton is directly correlated to the incoming proton energy and the spallation source. [4] Regardless of the spallation material, proton beams should be designed to operate at the energy level that maximizes the neutron yield per proton per eV, as this will yield the greatest number of neutrons, and thus most amount of fission, for the same energy input. Neutron yields are below 10 for most materials with incident proton energies at 500 MeV and saturate around 2 GeV, with the peak of neutron yield per proton per eV lying between these two extremes. [4]
When choosing the appropriate material, additional factors must be evaluated. Heavier elements, such as natural Uranium and Thorium, produce a higher number of neutrons from spallation, with Uranium producing nearly 40 neutrons per incident proton at 1.5 GeV. [4] In comparison lead will produce 20 neutrons per incident proton at that same energy level. [4] From a neutron economy perspective, heavier elements are far better neutron emitters than lighter elements.
However, when considering challenges of heat deposition and waste generation, heavier elements pose a challenge. The principal one lies in the byproduct of uranium spallation. Upon collision with the high energy protons, most uranium undergoes fission and produces large volumes of minor actinides and transuranics. [4] The generation of the waste significantly complicates the reactor's second objective of incinerating (through transmutation) nuclear waste.
Lead, on the other hand, presents particularly attractive characteristics due its waste products with Lead-Bismuth Eutectic (LBE) also showing an advantage over Uranium and Thorium, despite the complicated production of radioactive Polonium from proton and neutron reactions with Bismuth. [4] This suggests that the use of lead or LBE as a spallation source can contribute to the overall goal of reducing radiotoxicity of nuclear waste. Its contribution to the generation of radioactive waste is far less than that of fissile materials such as Uranium - however at a cost of generating half the number of neutrons. [4]
The first design proposed by Carlo Rubbia in 1993 was the Thermal Energy Amplifier. The reactor was similar to the design discussed above, operating with a combined LINAC and cyclotron to accelerate protons to 1.5 GeV at an incoming current of 10 mA. [1] The under-moderated water ADS design used Thorium as a fuel and spallation target, generating an estimated 40 neutrons per incoming proton. [1,4] The design operated at the thermal neutron spectrum, using water to moderate neutrons to the thermal level. [1] The design operated with less water than conventional critical reactors (under moderation), resulting in a higher fraction of fast neutrons while minimizing neutron loss to water by capture (reduced from 20% existent in most PWRs to less than 2%). This enabled efficient Thorium breeding in the reactor. [1]
While the T-EA could theoretically operate without using enriched fuel, the costs of initially irradiating Thorium with neutrons from the spallation source to generate enough U-233 to kickstart subcritical fission would make the reactor economically unviable. [1] Thus, Rubbia proposed an initial fuel loading of highly enriched U-235, either directly dissolved into the Th-232 fuel or contained in some auxiliary elements. [1] A suggested 1% of U-235 was to be be mixed into the Thorium to kickstart the reaction. Once kickstarted, the Thorium would breed U-233 to an equilibrium level of 1.35%. This would keep combustion ongoing for several years until poisoning of the bar by the fission products required fuel replacement. [1]
However, two years after initially proposing the T-EA design Rubbia moved to update his design with a new proposal called the F-EA, which he claimed was far superior to his initial proposal. [2] Firstly, the new design offered a far greater burn-up, generating up to 100 GW-day of energy per ton of Thorium burnt. This allowed for far more energy to be generated from fuel consumed compared to the T-EA, which offered 50 GW-day of energy per ton of Thorium burnt. [2] More energy per ton of fuel input could theoretically translate to a more economically efficient reactor if other costs remained the same, as the design would spend far less on fuel for the same energy output. The greater burn-up was made possible by a reduced capture cross rate by fission fragments for fast neutrons compared to thermal neutrons, resulting in less poison accumulation. [2] Neutron capture cross sections are much smaller both for fission fragments and for newly produced Uranium, Protactinium and Neptunium isotopes with the fast spectrum.
Beyond a greater burn-up, the design offered a far greater energetic gain (ratio of the total energy output from the device per energy deposited by the high energy proton beam). It would reach a factor of 100 to 150, greatly exceeding the gain estimated for the T-EA design, which averaged between 20 and 40. [1,2] This would mean that for the same MWh of energy deposited by the beam, the F-EA could offer 2.5 to 7.5x more energy output from the reactor than the T-EA. The higher gain is due both to a more efficient energy target configuration and to a larger practical value of neutron multiplication. [2]
As a fast breeder rather than a thermal breeder, the F-EA is significantly different from the T-EA. Firstly it operates without a moderator to allow fast neutrons to predominate. To operate without a moderator it must also operate without a water as a coolant, and must thus have an alternative coolant. Critical fast breeders have historically operated with either Sodium or Lead (or Lead-Bismuth eutectic) as a liquid metal coolant. [2] Safety problems and operational challenges experienced with Sodium fast breeders pushed Rubbia away from the Sodium alternative. [23] Furthermore, the possibility of using Lead as a coolant and as a spallation target, due to its relatively high neutron yield per proton prompted Rubbia to push forward with a Lead design. [2] An alternative design with LBE could be used, offering an advantage of lower melting temperatures but at the cost of generating more radioactive waste, as outlined above. [4] Both coolant options offer an additional advantage to the under-moderated T-EA design in operating at slightly higher temperatures without generating significant steam pressures, allowing for an efficiency gain in comparison to the T-EA design.
When proposing the F-EA Rubbia followed design of the Superphénix Sodium Cooled Fast Reactor in France with small modifications to accommodate the needs of an accelerator-driven system. [5] The F-EA is a pool-type reactor with its core, target unit and primary heat exchanger submerged inside a pool of liquid lead coolant. [2] The design operates with a lower proton energy beam of 1.0 GeV, as the neutron yield per incident proton per eV is maximized for Lead at this level. [4,5] The proposed proton current is higher at 12.5 mA and is originated from a collection of modular cyclotrons connected to an initial LINAC. [5] The design was planned to generate of 1500 MWth of energy and 675 MWe of primary electrical power, reflecting a thermodynamical efficiency of 45%. The improved efficiency was primarily due to the higher operating temperature of 550°C to 600°C enabled by the use of a Lead coolant. The expected energetic gain from the initial design was 120, within the 100 to 150 range expected. [5]
In light of the possible advantages from the F-EA design, a demonstration facility was proposed known as the experimental accelerator drive system (XADS). [6] The demonstrator proposed by Ansaldo (also known as the LBE-XADS) followed the F-EA design but used a Lead-Bismuth eutectic coolant (LBE) rather than a Lead coolant. [6] As the objective of the experiment was to evaluate the technical feasibility of the F-EA rather than incinerate nuclear waste, LBE was a more suitable combined coolant and spallation target, due to its lower melting point compared to molten Lead. [4]
 |
| Fig. 2: Overview of ADS Reactor Vessel. (1) Proton beam inside target unit. (2) Target unit with LBE spallation source (removable). (3) Enriched fuel for fission. (4) Dummy elements. (5) primary coolant lBE (which is separated from the LBE in the target unit to avoid contamination). (6) Heat exchanger. [10] (Image Source: F. Rabbat Neto) |
Beyond changing the coolant and spallation target, the demonstration facility proposed to use an existing and approved fuel instead of Thorium-based fuel, thus avoiding the additional certifications needed for Thorium fuel. [6] The suggested fuel was the well known and certified Supherphenix-type uranium-plutonium MOX fuel (SPX). [6] This fuel consists of 82% natural Uranium and 18% Plutonium containing some traces of both Americium and Neptunium. 71% of the Plutonium was fissile U-239, leading to an enrichment of 12.78% (0.18 × 71% = 12.78%). Fig. 2 shows the main components of the design.
The energy generation process begins in the Target Unit (part 2). The high-energy high-flux proton beam (600 MeV and 3.23 mA) generated in a combined LINAC and cyclotron accelerator is directed by the proton beam pipe (part 1) through the middle of the Target Unit. It then collides deep down with the liquid LBE flowing inside the Target Unit. [6] Upon collision the LBE undergoes an intranuclear cascade, ejecting on average 15.04 neutrons per incident proton. These neutrons expand outwards at high energies to the surrounding enriched subcritical fuel core (part 3). [6] The fast neutrons collide with the fuel, transmuting fertile fuel into fissile fuel and causing fission of the fissile material. The fission generates heat, which is collected by the LBE primary coolant and transferred by convection upwards to the heat exchangers. [6] The heat exchangers transfer the heat to power a steam generator, generating energy in a similar manner to all thermal power plants.
It is worth noting that the target unit (parts 1 and 2) is an entirely removable component that can be replaced as needed during operation of the reactor. This also means that the LBE spallation source is physically separated from the LBE primary coolant to avoid radioactive contamination of the primary coolant, a further modification from the original F-EA. [6]
Rubbia's work provided some interesting insights about the theoretical energy output and operating conditions of the proposed plant design. The original design conceived was an 80 MWth plant, with 1.94 MWe proton beam (3.23 mA × 600 MeV = 1.94 MW). [6] This represents an energetic gain of approximately 40 (80 MWth/1.94 MWe = 41.3). This is significantly below Rubia's original proposed gain of 120 for the F-EA but greater than the energetic gain estimated for the T-EA. In the presented design, every neutron from spallation would be generating on average 26.7 neutrons from fission. The energy output would be the same as that of a commercial LWR. Considering the intricacies of the design, with added components such as an accelerator and a spallation target, questions on costs immediately arise, as the design is using the same energy source to result in the same energy output, with a far more complex and expensive system. Nonetheless, questions of neutron economy, fuel economy and operating temperatures must be considered to better evaluate the energy output and efficiency of the design.
Following the Superphénix design, the experimental reactor was projected to operate at 400°C, significantly below the suggested temperatures for the F-EA. [2,6] While the main F-EA design could achieve some gain from operating at higher temperatures, it is worth noting that 600°C lies above stainless steel creep temperatures (~566°C) and thus material challenges could arise that could result in higher costs and eat into the gain from greater efficiencies. [6] In the experimental design, the proposed operating temperatures are comparable to the operating temperatures of LWRs and suggest that the efficiency of this reactor would at best be equivalent to the efficiency of LWRs, around 30%. [7] Using the efficiency of LWRs as an estimate for the efficiency of the ADS we can calculate that 24 MWe would be generated from 80 MWth output.
However, it is important to remember that the 80 MWth requires a continuous supply of high energy protons to power spallation and these protons require energy to be accelerated to 600 MeV at a 3.23 mA flux, as proposed in Rubbia's design.
If the accelerator were 100% efficient then 1.94 MWe of the electricity produced would have to be fed back to the accelerator to power the spallation. Unfortunately, accelerators are far from 100% efficient, with estimates placing grid-to-beam efficiencies at 18.1%. [8] This would mean that the 1.94 MW beam would consume a total electric power of
| 1.94 × 106 Watts 0.181 |
= | 10.72 × 106 Watts |
Since the reactor's nominal electric output would be 24 MWe, 45% of the reactor's electricity production would have to be fed back to the accelerator. This would leave a remaining 13.29 MWe of energy to be supplied to the grid, reducing the overall efficiency of the system to 16.62%. This has very important economic considerations, for if the reactor were to cost the same per thermal watt as a LWR, its cost per electric watt would be triple. Even if the reactor were able to operate at 45% efficiency as proposed by Rubbia in the F-EA design, the loss from the accelerator would be significant. Considering a net energy output of 36 MWe (80 MWth × 0.45 = 36 MWe), with 10.72 MWe required to power the accelerator only 25.29 MWe of power would be left to supply to the grid resulting in an efficiency of 31.12%, on par with the 30% efficiency of LWRs. However, this efficiency equivalence would come with the added costs of replacing stainless steel pipes with materials expensive nickel steel alloys to avoid creep at such high temperatures. Given that accelerators generating 600 MeV and 3.23 mA proton beams are not likely to be cheap, the overall costs of the system could make it an unviable energy generation source.
While energy generation in an ADS may present significant challenges for limited increased safety benefits resulting from subcriticality, the reactor's external fast neutron source allows for the use of fuel with suboptimal neutronic properties, making it a possible candidate for waste transmutation. Transmutation is a achieved by neutrons colliding with nuclei and causing capture or fission. Capture can result in instability and a beta decay of the nuclei, transmuting the element into another element possibly nonradioactive or with far shorter half-life. Fission can break down the element into smaller particles that could be nonradioactive or have a shorter half-life. The ultimate objective for waste transmutation is to reduce the concentrations of radioactive isotopes or isotopes with long half-lives so that the end transmuted waste can be stored for a far shorter period, in a significantly smaller volume, until it is no longer a radiation danger.
As transmutation relies on the supply of neutrons, a question arises as to why ADS subcritical reactors are better fit to deal with transmutation compared to thermal LWRs and fast reactors. At thermal neutron energies, minor actinides, such as Np-237 Am-241, have too low fission and capture cross-sections to undergo fission and sustain a critical reaction. [9] Even if highly enriched fuel were used to increase neutron flux and maintain the reaction, the fission and capture cross-section are such that more neutrons would be captured than would cause fission. This would result in more transuranic and minor actinides waste being produced than consumed, thus failing to reduce transuranic and minor actinide waste. [9]
In fast reactors, higher neutron fluxes emitted at the fast energy spectrum provide a significant opportunity to transmute minor actinides and transuranic elements. [9] The epithermal and fast neutron spectrum significantly increases the fission cross-section of Np-237 and Am-241, allowing these elements to undergo fission, thus providing enough neutrons for criticality to be maintained at lower enrichment concentrations. [9] However, in critical systems, higher concentrations of transuranics and minor actinides comes with a significant safety concern. When minor actinides and transuranics are doped into the system and operate as fissile material they produce significantly fewer delayed neutrons compared to U-235. [4] In critical systems delayed neutrons provide time for operators to adjust control rods to sustain the reaction at criticality. Reduction in the fraction of delayed neutrons reduces the time between the initiation of fission and continuation of the chain reaction, ultimately making it extremely difficult to control a critical core, putting the core at risk of supercriticality. [4]
On the other hand, in a subcritical fast reactor, delayed neutrons do not play a role in controlling the reaction, as the reaction sustained by the neutrons from spallation rather than critical fission. [4] As a result, the reactor is not being actively controlled to manage its neutron population at criticality. Thus, the ADS would benefit from the fast neutron energy levels of fast reactors (maximizing fission to capture ratio of transuranics and minor actinides), without adding any significant safety concerns from lowered fraction of delayed neutrons. The neutrons captured would transmute the transuranics and minor actinides into other elements while the fission would break them down into smaller daughter products. [4] Having spallation drive the reaction and fission augment the neutron flux results in a high neutron flux that could ultimately incinerate minor actinides and transuranic elements without posing supercriticality concerns. If a net reduction in the transuranic elements and minor actinides is achieved, Accelerator Driven Systems would prove the sole theoretically feasible and safe alternative for transuranic and minor actinide transmutation and incineration.
The ADS design also permits for transmutation of other fission products that contribute significantly to the long-term toxicity of radioactive waste: long lived fission fragments (LLFFs) such as Tc-99 and I-129. These products have high capture cross sections in the epithermal regions and consequently could not be transmuted in thermal reactors such as LWRs where thermal flux dominate. [4] In an ADS and other fast reactors, LLFFs can be placed around the fuel core in an area known as the incineration blanket where neutron flux is particularly high at the epihthermal region. [6] Neutrons generated in the fuel core have collided with heavy metals and slowed down incrementally, arriving at the incineration blanket at the epithermal level. [6] At the incineration blanket the neutrons collide with larger elements remaining at the epithermal region and being captured by the LLFFs, transmuting them into nonradioactive isotopes or isotopes with shorter half-lives. [6]
When analyzing waste incineration capabilities, a key measure to first consider is the efficiency of the fuel, meaning how much energy is extracted per ton of fuel burnt. In regularly PWRs with UOX fuel at LEU enrichment levels, burn-ups were around 33 GW days per tonne of heavy metal incinerated (GWd/tHM). This meant that for every tonne of fuel burnt, 33 GWd (equivalent to 792 GWh) of energy was generated. PWR using more enriched MOX fuel can achieve higher burn-ups, as more enriched material can burn for longer and generate more energy, resulting in an estimated burn-up of 43.5 GWd/tHM. [6] For the F-EA, Rubbia estimated burn-ups of over 100 GWd/tHM with Th-232 fuel, yet with the XADS the burn-up was estimated at 23 GWd/tHM with the SPX fuel. [2,6] The burn-up for the F-EA is significantly better than the burn-up from PWRs with UOX and MOX fuel, suggesting an extensive fuel economy in the adoption of F-EA. While this doesn't directly contribute to waste incineration, it is nonetheless important because it shows how less fuel is used to produce the same amount of energy, thus reducing the waste fuel inventory.
The lower burn-up for the XADS has a number of causes. But since the reactor was designed to test the viability of an ADS facility and waste incineration, it is difficult to compare its burn-up directly to the burn-up of commercial reactors. Given that the burn-up of fuels are different and the type of plants are different, the numerical comparisons between the simulated waste generation of the three designs listed in Table 1 below (PWR with UOX, PWR with MOX and XADS with SPX) is of minimal importance. The important consideration is whether kilograms of waste are generated (positive) or incinerated (negative) per tonne of fuel burnt.
|
||||||||||||||||||||||||||||
| Table 1: Kilograms of TRUs and LLFFs produced at end-of-cycle in PWRs and XADS per tonne of spent fuel. [6] |
It may be seen from Table 1 that, operating with the existing fuel that was used to power the Superphénix reactor (18% Pu-239 and 82% U-238), the XADS had a net reduction in Plutonium fuel and in LLFFs with a minimum increase in minor actinides and other transuranics in comparison to light water reactors. This demonstrates a particular advantage of the reactor in comparison to LWRs, as they can serve as an theoretically possible way of reducing Plutonium and LLFFs stockpile in waste fuel.
However, the use of SPX fuel in the XADS shows a production of minor actinides, suggesting that regular operation of the XADS alone is not a feasible means to effectively eliminate waste. Introduction of specific waste elements and an alternate fuel choice must be evaluated to effectively deal with the waste.
In light of the increased radiotoxicity of SPX fuel upon irradiation, the XADS was evaluated under its original fuel design of Thorium fuel doped with a fissile element for reaction kickstart, such as Pu-239. Th-232 produces far fewer TRUs and MAs than MOX fuel because it lies 5 neutron captures away from TRUs. This ensures that it can work in a mode that destroys more TRUs than it produces. [4]
|
|||||||||||||||||||||
| Table 2: Kilograms of TRUs and LLFFs produced at end-of-cycle in XADS per tonne of spent fuel. [6] |
Table 2 compares the waste generated/incinerated by the XADS when using SPX fuel vs ThPuO2 fuel. It may be seen that the use of ThPuO2 fuel can significantly reduce TRUs stockpile with small increases in MAs, except for Americium. It should also be noted that with Th-232 fuel, LLFFs such as Tc-99 are also decreased as they transmuted into other radioactive isotopes with shorter half-lives. [6] This would suggest that doping fuel with TRUs and doping the incineration blanket with LLFFs, it is possible to achieve a net reduction of TRUs and LLFFs with a small increase in MAs.
It is interesting to evaluate the TRU elimination in terms of energy output. The XADs may eliminate ~28.6 kg/TWh of Plutonium, ~2.2 times that produced in a PWR using UOX (~13 kg/TWh). [6] The transmutation of large quantities of LLFFs is also possible without disturbing the multiplication of the system, posing fewer problems in terms of safety and neutron economy. [6]
Effectively, these numbers imply that adopting an F-EA could eliminate continuously the TRUs and LLFFs waste generated by our existing LWRs, thus potentially allowing a progressive shift from the Uranium-Plutonium fuel cycle to the less polluting Thorium fuel cycle operated in fast ADS.
To dope Th-232 fuel with TRUs and LLFFs these materials must be partitioned from the remainder of the fuel rods. It is true that in some countries the PUREX process already separates Plutonium and Uranium from the remainder of the fuel, but further work would be needed for the separation of the long-lived fission fragments. [4] Furthermore, given the different neutron spectra required to transmute transuranics compared to LLFFs, this group of elements must be separated and concentrated into fuel matrices. This additional layer of reprocessing and portioning can pose significant economic challenges for the deployment of ADS as an effective waste transmutation strategy. In countries such as the USA, PUREX reprocessing isn't even currently implemented, meaning that the entire reprocessing infrastructure would have to be built from scratch, adding significant financial and technical hurdles for effective implementation.
Accelerator Driven Systems represent a promising next-generation nuclear technology that could help address key challenges facing the nuclear industry, namely safety concerns with conventional reactors and the pressing issue of nuclear waste disposal. By utilizing a subcritical core sustained by an external neutron source from spallation, ADS designs such as the F-EA offer enhanced safety features compared to critical reactors. However, the real potential of ADS lies in its waste transmutation and incineration capabilities.
The fast neutron spectrum and high neutron flux in an ADS allows it to effectively transmute problematic transuranic elements and long-lived fission fragments. Thorium-based ADS designs optimized for waste burning could drastically reduce the radiotoxicity of nuclear waste. The buildup of minor actinides would nonetheless still require alternative incineration solutions or long-term geological repositories. An ADS could potentially eliminate most of the high-level waste from multiple light water reactors.
However, despite the technical merits, ADS still faces significant challenges. The technology has yet to be demonstrated on a experimental or commercial scale, with many design aspects still requiring validation. The complex partitioning processes needed to separate waste elements add additional technical hurdles. Most importantly, the economics of ADS remain uncertain. The addition of a high-power accelerator and spallation target introduces substantial costs compared to conventional reactors. The inherently lower thermal efficiency due to the accelerator power requirements further hinders the economic competitiveness of ADS for power generation.
Ultimately, for ADS to be viable as a waste disposal solution, it must prove more cost-effective than directly disposing of spent fuel in a geological repository. Competing against "a hole in the ground" sets a very high bar for the allowable costs of ADS. Significant technological progress and economic optimization will be needed for ADS to fulfill its envisioned role in the nuclear fuel cycle. Nonetheless, given the immense challenges posed by nuclear waste, ADS remains an important avenue of research and development in the quest for a more sustainable nuclear energy future.
© Freddy Rabbat Neto. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
[1] F. Carminati et al., "An Energy Amplifier for Cleaner and Inexhaustible Nuclear Energy Production Driven by a Particle Beam Accelerator," European Organiation for Nuclear Research, CERN, CERN/AT/93-47, November 1993.
[2] C. Rubbia "A High Gain Energy Amplifier Operated With Fast Neutrons," AIP Conf. Proc. 346, 44 (1995).
[3] A. Herrera-Martínez, Y. Kodi and G. Parks, "Transmutation of Nuclear Waste in Accelerator-Driven Systems: Thermal Spectrum," Ann. Nucl. Energy 34, 550 (2007).
[4] Y. Kadi and J. Revol, "Design of an Accelerator-Driven System for the Destruction of Nuclear Waste," European Organization for Nuclear Research, CERN, September 2001.
[5] C. Rubbia et al., "Conceptual Design of a Fast Neutron Operated High Power Energy Amplifier," European Organization for Nuclear Research, CERN, CERN/AT/95-44, September 1995.
[6] A. Herrera-Martínez et al., "Transmutation of Nuclear Waste in Accelerator-Driven Systems: Fast Spectrum," Ann. Nucl. Energy 34, 564 (2007).
[7] "The Future of Nuclear Energy in a Carbon-Constrained World," Massachusetts Institute of Technology, 2018.
[8] A. Kovach et al., "Energy Efficiency and Saving Potential Analysis of the High Intensity Proton Accelerator HIPA at PSI," J. Phys. Conf. Ser. 874, 012058 (2017).
[9] C. D. Bowman et al., "Nuclear Energy Generation and Waste Transmutation Using an Accelerator-Driven Intense Thermal Neutron Source," 1992, Nucl. Instrum. Methods Phys. Res. A 320, 336 (1992).
[10] P. Baxevanis, "Accelerator Driven Subcritical Reactors," Physics 241, Stanford University, Winter 2013.